Accession Number : ADA186386

Title :   Estimation of Multivariate Binary Density Using Orthonormal Functions.

Descriptive Note : Technical rept.,

Corporate Author : PITTSBURGH UNIV PA CENTER FOR MULTIVARIATE ANALYSIS

Personal Author(s) : Chen, X R ; Krishnaiah, P R ; Liang, W Q

PDF Url : ADA186386

Report Date : Dec 1986

Pagination or Media Count : 17

Abstract : In a number of situations, the experimenter is confronted with the statistical analysis of the data which is binary in nature. For example, one may be interested in diagnosis of the disease on the basis of symptoms. The reliability of complicated systems can be studied by examining as to whether its components are functioning or not. In image processing, a picture is classified on the basis of two grey levels like white and black using some threshold value. We may assign a score of 1 or 0 according as the grey level is white or black respectively. So, it is important to study the problems of estimation of multivariate binary density. Cencov expressed continuous multivariate density as a series of orthonormal functions. Bahadur expressed the multivariate binary density as a series. Ott and Kronmal expressed the density as a series involving Walsh functions. Liang and Krishnaiah also expressed the density in terms of Walsh functions but the coefficients in their series are different from those used by Ott and Kronmal. This paper is a continuation of the work done by Liang and Krishnaiah.

Descriptors :   *DENSITY, *MULTIVARIATE ANALYSIS, *THRESHOLD EFFECTS, *WALSH FUNCTIONS, *ESTIMATES, COEFFICIENTS, DIAGNOSIS(GENERAL), DISEASES, IMAGE PROCESSING, RELIABILITY, SIGNS AND SYMPTOMS, STATISTICAL ANALYSIS

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE